So instead of 125Meg to run a basic wordprocessor it will now be 500Meg plus..
I've used StarOffice / OpenOffice / LibraOffice for more than 20 years because it mostly works, it reads all .doc files (which MS Word doesnt) and is unlikely to go away. No other reason.
The codebase was always a mess and there is absolutely no need for a 125Meg footprint for a level of functionality little different than what took a around a 1Meg footprint back in the late 1980's. Just very sloppy coding. Of course we had to write the core line layout and frame layout functions in assembler back then (for speed) but as I look at the feature set, when it comes to actual real world use scenarios, I just see massive bloat.
If anyone bothered to actually profile the code they would find that a good 95% plus of that 125M was due to very bad code partitioning with serious dependency fan out. Low frequency calls accounting for most of the module inter-dependency. And a good 80% of the feature code used to support user work scenarios that account for < 0.001% of all actual real world usage.
Featuritis makes perfect sense for commercial products in a regular upgrade cycle. You need the income. But for not-commercial product featuritis just adds to code entropy with zero change in the feature use frequency profile. Purely dead weight.
So Libre Office is usable for casual use but for real world daily professional use it was never really in the running. I see far too many real bugs that indicate serious code problems even in causal use. When you have written a couple of WP/DTP codebases over the decades its very easy to not only spot the bugs but work out the most probably why for them.
So why dont I dive in and fix the bugs I see? Because I looked through the source code when Sun first open sourced it and having waded through the swamp for a little while said - No Thanks., I'm outta here. Seen way too much code like in the past. In commercial products. Always either thrown away eventually and rewritten or else the product died eventually under the weight of bugs. I need to get paid good money to do that kind of very frustrating low productivity work
So any WebAssembly version will be unstable and buggy as hell. Absolutely guaranteed.
Will look good on someones resume though.